Reliability and validity of data obtained from alcohol, cannabis, and gambling populations on Amazon’s Mechanical Turk.
Bibliographic record
Abstract
Researchers recently have begun using Mechanical Turk (MTurk), an online crowdsourcing platform, to recruit addiction populations. However, whether the data obtained from substance users and gamblers on MTurk are reliable and valid is unknown. Herein, we assessed the internal and retest reliability of and concurrent and convergent validity of data obtained from addiction populations on MTurk. Current drinkers (N = 208), cannabis users (N = 200), and gamblers (N = 200) residing in the United States completed measures of alcohol, cannabis, and gambling severity, psychological constructs (e.g., impulsivity) related to addictions, overt and subtle measures of valid responding, and motivations for completing MTurk studies. Of the original sample, 88-92% of participants who provided informed consent for recontact completed a reassessment 1 week later. The internal consistency of the addiction severity measures ranged from α = .75 to .93. The stability over 1 week ranged from κ = .57 to .70 for categorical classification, and intraclass correlation coefficient (ICC) = .71 to .86 for continuous measures. The addiction measures were significantly correlated with each other and with other constructs related to addictive behaviors. Overall, 80-85% of participants provided valid responses. They reported attending and answering questions honestly, with financial motives being the most frequently endorsed motivation. After invalid responses were excluded, results remained the same for alcohol and gambling, but significant differences emerged for the cannabis sample. The results suggest that the self-report data obtained from alcohol and gambling populations are of high quality, however, caution is warranted with cannabis populations. MTurk shows promise as a recruitment tool for some addictive behaviors. (PsycINFO Database Record
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.028 | 0.068 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".